Supply chain project failure is not uncommon. These projects are inherently complicated and are by nature high-risk. I watched Top Gun: Maverick starring Tom Cruise this weekend, so you’ll notice my fighter-pilot analogies throughout this piece (apologies in advance). If you don’t hit the target on supply chain projects, there is a distinct possibility that you will crash and burn. So, to help you avoid having to reach for the ejection-seat handles, I’ve compiled a pre-flight checklist (later in this post) for your consideration.
Having led hundreds of supply chain projects over the course of more than twenty years of consulting at OPSdesign, I’m pleased to say that we’ve never been involved in a “crash and burn” scenario, other than those we’ve been asked to clean up after a catastrophic failure caused by others.
If you are lucky, a failure may only be an unachieved service level mark, underperforming productivity metrics, or failure to achieve the anticipated return-on-investment (the equivalent of missing your target, having a piece of your wing shot off, but still making it back to the aircraft carrier with only a few flesh wounds). You’re alive, but you have a lot of explaining to do!
If the project really goes badly (and we’ve been asked to clean up the mess on quite a few), you might find yourself in a shiny new 500,000 sq. ft. e-commerce distribution facility having spent tens of millions of dollars, only to find that processes, systems, infrastructure, and labor elements have not been fully baked; integration has failed; the project has been poorly planned and managed; employees have not been properly trained; you are four months behind schedule; and that you simply cannot ship an order. The board wants to know what went wrong and who is responsible? CRASH AND BURN!
PRE-FLIGHT CHECKLIST (why supply chain projects crash and burn):
( ) Underestimating complexity and timeline: We routinely (and respectfully) walk away from project opportunities that have been architected with unrealistic, high-risk, and unachievable timelines. If you only have 6 months to map processes, select, and implement a new WMS; or 9 months to design, fabricate, install, train, test, and debug a new distribution center, sorry – too late! It is not uncommon for these projects not to get the attention they deserve and the necessary funding until the “operational pain” has become unbearable. By that time, the window of opportunity to strike may have closed. This is the number one reason for a crash and burn project. Critical path elements, tasks, and milestones must be mapped at the onset of supply chain initiatives to rationally define project duration (balancing risk tolerance). Shoehorning a mission into an insufficient available timeframe increases the likelihood of failure. Poor planning, flawed assumptions, unrealistic schedule, and fuzzy critical path/milestones compound this issue.
( ) No clear and empowered project champion (internal): Who owns this project, can cut through red-tape, and align the various departmental interests (ally the disparate silos to a common goal)? Without this leader, politics and departmental infighting make for a messy battlefield. It can be impossible to define and achieve a successful outcome without a strong project champion willing and able to lead in the war-room.
( ) No clear and empowered project execution leaders (internal/external): While the project champion may have architected the strategic battle plan, he or she is not going into combat. The tactical execution leaders may be entirely internal or some combination of internal and external (supply chain consultants/interim project managers) resources. In the case of internal leadership only, it is imperative that he/she has successfully executed similar initiatives in recent history. Far too often, supply chain executives/operators get cocooned in one position or operation for many years and simply are unable to cross-pollinate best practices and successfully manage these complex projects while attending to their existing daily operational responsibilities. Operators become good at operating. There’s a difference between making the donuts and designing the donut factory. Does the internal execution leader have the experience, subject matter expertise, and horsepower, to go it alone?
( ) Inadequate analysis, Ill-defined business case: Most supply chain design/optimization initiatives begin with historical information (old data), multiplied by forecast data (bad information) to determine design year criteria. Taking this suspect data and compounding it (typically) 5-years forward is a shaky foundation upon which to build your initiative. To reduce risk, the volatile model elements must be sensitivity tested severally and in aggregate. What if our 2% year over year SKU proliferation estimate is understated? What if our inventory turn target is not achieved? What if we add a product family that is different from our current cube and weight model? What if our customers require additional value-added-services we currently do not provide? Rest assured your business will change over time and your history and forecasts will be increasingly inaccurate as each year is ticked off the anticipated lifespan of your supply chain project. Prudent sensitivity testing can help define and avoid unnecessary risk.
( ) Failure to compare all viable alternatives to the base model: The base model may be the current operations or network (the “do nothing” alternative) or the lowest risk, lowest capital approach when something must be done (as in the case of entering a new market channel such as e-commerce). Before committing to battle, you’ve got to run war-games (modeling or simulation) to predict the outcome. A thorough engineering-based comparative analysis of processes, systems, infrastructure, and labor alternatives must be performed. Each alternative must be quantified/qualified in terms of capital cost, recurring cost (labor), capacity, throughput, service impact, flexibility, scalability, and risk. Just because a vendor of a particular solution (goods-to-person AS/RS or shuttle systems, autonomous mobile robots, or voice-picking technology, etc.) demonstrates a favorable ROI vs. your current operation, doesn’t mean it’s the best choice. How does that solution stack up against all viable alternatives?
( ) Inability to pull the trigger, make a decision, weigh risk/reward: Otherwise known as “analysis paralysis”. Geographic network modeling analysis and design of warehousing, distribution, and e-commerce fulfillment operations require a minimum of 12-16 weeks to do the necessary, discovery, strategy, and planning before any implementation steps can begin. Exercising due diligence is a prerequisite for success; however, at some point hard decisions (based on the science and art in the first few months of the project) must be made. These decisions are informed by hundreds of hours of engineering work, but the risk vs. reward (to fly or not to fly the mission and what that mission looks like) is ultimately up to the project champion and the board. Someone needs to give the order to catapult off the deck. Sitting on this decision-making matrix too long can adversely compress the remaining time in which to complete the mission successfully.
( ) Poor execution and project management against the plan: Even the most thoughtful and well architected battle plan needs to be expertly executed in the field. Frequently, a solid plan fails due to lack of project management experience or the needed horsepower to bring the project to the finish line. There will be challenges, opportunities, and unforeseen conflicts along the way. A cohesive project team led by an experienced implementation manager can circumvent skirmishes, avoid ambushes, and still reach the target effectively. Communication, command, and control are key elements to his or her success, just as they are in military operations (up and down the chain of command).
( ) Lack of contingency planning and risk mitigation measures: War is messy and the plan of attack on your supply chain initiative will need to be agile and resilient. Too many supply chain operations optimization projects fail simply due to lack of a solid contingency plan and risk management protocol. One crash and burn project we were asked to take over from another consulting firm comes to mind as the perfect storm for failure. It was a consolidation of an East and West Coast warehouse to one very large Central facility. Leases on both existing facilities were set to expire shortly before the targeted go-live at the new consolidated location (no viable contingency in place if the new facility was not ready – the old ones had to be vacated by lease expiry). Compounding that major risk was the fact that very few employees from the existing operations were going to move to the new city and staff the new operation. There was insufficient time allotted to recruit and train new associates. In addition to the high levels of material handling automation/mechanization, a new Enterprise Resource Planning system (ERP), new Warehouse Management System (WMS), new Warehouse Control System (WCS), and new Human-Machine Interface (HMI) required the integration of all processes and disparate data systems. While all projects carry inherent risk, this was simply flying low-level through the fog at Mach-speed without instruments. The crash was ugly and inevitable. The client’s C-suite and investors wanted to know who was responsible. Not a good position to find yourself in! Identify risk factors, determine cost/risk tolerance parameters, and define actionable contingency plans. Review and update the risk avoidance and contingency plans throughout the project so that they may be executed quickly when a guided missile comes your way. There will not be time to think and devise a plan under the heat of battle.
( ) Poor pre-launch stress-testing/debugging: The branches of the military run drills, simulations, war-games etc. to test their plan and instill muscle-memory. Prior to go-live, all processes, systems, infrastructure, and labor deployments must be stress-tested to assure that mission failure does not occur at go-live and beyond. Stress testing plans must be designed and rolled-out to emulate peak data loads, check interfaces, and mimic physical throughput requirements on physical infrastructure, mechanization, and automation elements. Finally, the human-machine interfaces must also be validated at peak demand throughput volumes. Try to break the system while there is time to address the failure before go-live. Finally, do not schedule 100 percent of peak levels on day one. Start with a reasonable volume and ramp up slowly over the first two weeks to uncover and address hidden issues.
( ) Post-launch follow-up on metrics/KPI achievement against the targeted goals: Finally, the war is not over until the battlefield is cleared and casualties accounted for. This means revisiting the originally targeted objectives (capacity, throughput, cycle time, accuracy, productivity metrics, KPIs, etc.) and tracking actual performance against the target. Where gaps exist, analyze the design criteria and the intended Standard Operating Procedures (SOPs) to determine why there is a shortfall. Take corrective action accordingly and report your success up the chain of command.
In closing, I apologize for all the trite military cliches and Top Gun analogies, but I guess I was a bit amped up after watching the film Friday night. Specific apologies for all the Navy references go out to three of my valued employees who are Veterans of the United States Army and graduates of the United States Military Academy West Point. Thank you for your service!
